Systemic administration of an RNA binding and cell-penetrating antibody targets therapeutic RNA to multiple mouse models of cancer

Abstract

There is intense interest in the advancement of RNAs as rationally designed therapeutic agents, especially in oncology, where a major focus is to use RNAs to stimulate pattern recognition receptors to leverage innate immune responses. However, the inability to selectively deliver therapeutic RNAs within target cells after intravenous administration now hinders the development of this type of treatment for cancer and other disorders. Here, we found that a tumor-targeting, cell-penetrating, and RNA binding monoclonal antibody, TMAB3, can form stable, noncovalent antibody/RNA complexes of a discrete size that mediate highly specific and functional delivery of RNAs into tumors. Using 3p-hpRNA, an agonist of the pattern recognition receptor retinoic acid-inducible gene-I (RIG-I), we observed robust antitumor efficacy of systemically administered TMAB3/3p-hpRNA complexes in mouse models of pancreatic cancer, medulloblastoma, and melanoma. In the KPC syngeneic, orthotopic pancreatic cancer model in immunocompetent mice, treatment with TMAB3/3p-hpRNA tripled animal survival, decreased tumor growth, and specifically targeted malignant cells, with a 1500-fold difference in RNA delivery into tumor cells versus nonmalignant cells within the tumor mass. Single-cell RNA sequencing (scRNA-seq) and flow cytometry demonstrated that TMAB3/3p-hpRNA treatment elicited a potent antitumoral immune response characterized by RIG-I activation and increased infiltration and activity of cytotoxic T cells. These studies established that TMAB3/RNA complexes can deliver RNA payloads specifically to hard-to-treat tumor cells to achieve antitumor efficacy, providing an antibody-based platform to advance the study of RNA therapies for the treatment of patients with cancer.

Fig. 1.. TMAB3 non-covalent binds to 3p-hpRNA and delivers it into cells to activate RIG-I.

(A) Electrostatic maps were generated for TMAB1 and TMAB3 with the D31N substitution in CDR1. (B and C) Quantification of TMAB1 and TMAB3 binding to single-stranded DNA (B) and single-stranded RNA (C) by ELISA. (D and E) Bio-layer interferometry measurements of TMAB3 binding of the 89-nt 3p-hpRNA (D) and for the 886-nt GFP mRNA (E) over time. Derived values for the dissociation constant (K_D) are indicated. (F) Hydrodynamic diameter measurement of TMAB3/3p-hpRNA compared to TMAB3 alone using dynamic light scattering. N=16 replicates. (G). Results of the RNA protection assay showing the effects of RNase on 3p-hpRNA alone or when complexed with TMAB3. (H) (left) Immunofluorescence staining of MC38 mouse colon cancer cells for TMAB3 (green), pre-treated with DMSO (top row) or the ENT2 inhibitors dipyridamole (middle) or NBMPR (bottom). Nuclei are stained blue with DAPI. (right) Quantification of TMAB3 nuclear fluorescence intensity per group, normalized to DAPI. Statistical analysis was performed using the student’s t-test for individual comparison. **** P < 0.0001. (I) Immunofluorescence staining of U2OS human osteosarcoma cells, where TMAB3 (red) and 3p-hpRNA (green) were labeled with different fluorescent tags, then combined to form complexes and added to the cells. Control samples contained an isotype control antibody instead of TMAB3. Nuclei are stained blue with DAPI (J) Quantification of type I IFN induction by TMAB3/3p-hpRNA and the indicated controls in THP-1 monocytes wild-type for RIG-I or with RIG-I knockout, both containing an IFN-response luciferase reporter construct. Data are mean ± SEM from n=4 replicates with statistical analysis performed by One-way ANOVA and multiple comparisons by Holm-Sidak **** P < 0.0001.

Fig. 2.. Intravenously administered TMAB3/3p-hpRNA targets tumors and distributes specifically to malignant cells in vivo.

(A) ENT2 mRNA expression in human pancreatic ductal adenocarcinomas TCGA (n = 185) relative to normal pancreatic tissue controls from GTex (n = 362). T-test **** P < 0.0001. (B) Experimental timeline for the biodistribution assay in Pdx1-Cre; LSL-Kras^G12D/+ mice (PanIN mice) and wild-type mice. Mice were treated intravenous with dye-labeled (14) or isotype control and harvested pancreata were imaged the next day. (C) IVIS imaging of harvested pancreata 24 hours after intravenous administration in PanIN and wild-type mice. (D) Experimental timeline for 3p-hpRNA delivery by TMAB3h in Pdx1-Cre; LSL-Kras^G12D/+ mice. (E) Relative amounts by RT-qPCR quantification of 3p-hpRNA presence in EPCAM⁺ (tumor) cells compared to CD45⁺ cells isolated from pancreata of treated mice. T-test **** P < 0.0001 (F) Experimental timeline for TMAB3 biodistribution assay in mice bearing orthotopic pancreatic KPC tumors. The tumors express luciferase and the antibody is fluorescently labeled for visualization. (G) IVIS imaging of the bioluminescent signal from luciferase-expressing orthotopic KPC tumor. (H) IVIS imaging of fluorescently labeled TMAB3 co-localized with the luciferase-expressing orthotopic KPC tumor. Mice were treated intravenously with 10 mg/kg of dye-labeled TMAB3 and processed for whole animal imaging by IVIS 24 hours later. (I) Timeline of the TMAB3h antibody biodistribution experiment in mice bearing HCT116 tumors. (J) Immunoblot assessment of TMAB3h in the indicated tissues using an antibody to the Fc portion of TMAB3h. Vinculin was used as a loading control. Quantification by Image J of band intensities is shown below the corresponding sample. (K) Timeline of the experiment assessing 3p-hpRNA delivery in mice bearing orthotopic KPC tumors. (L) RT-qPCR quantification of 3p-hpRNA delivery to EPCAM⁺ (tumor) cells compared to CD45⁺ cells isolated from orthotopic KPC tumors in treated mice. T-test, **** P < 0.0001. (M) Timeline of the TMAB3h antibody biodistribution in mice bearing HCT116 flank tumors. (N) Ex vivo IVIS imaging of harvested HCT116 tumors 24 hours after injection. (O) Quantification of antibody distribution in HCT116 flank tumors by IVIS imaging. T-test, ** P < 0.01. (P) Timeline of the 3p-hpRNA biodistribution experiment in mice bearing KPC flank tumors. Mice were treated with intravenous TMAB3/3p-hpRNA complexes in which 3p-hpRNA was fluorescently tagged or with tagged 3p-hpRNA alone. (Q) Ex vivo IVIS imaging of 3p-hpRNA in harvested KPC tumors. (R) Quantification of 3p-hpRNA distribution in KPC tumors by IVIS imaging. T-test, * P < 0.05. (S) Timeline of the antibody biodistribution experiment in mice bearing CT26 flank tumors, pre-treated with dipyridamole and then fluorescent TMAB3. (T) Ex vivo IVIS imaging of fluorescent TMAB3 in harvested CT26 tumors. (U) Quantification of TMAB3 biodistribution in CT26 flank tumors by IVIS imaging. T-test, * P < 0.05.

Fig. 3.. Intravenously administered TMAB3/3p-hpRNA complexes enhance survival and suppress tumor growth in orthotopic pancreatic cancer and medulloblastoma mouse models.

(A) Timeline for the efficacy experiment in mice bearing orthotopic KPC tumors. Mice were treated with TMAB3/3p-hpRNA, gemcitabine, TMAB3/3p-hpRNA plus gemcitabine, or with PBS, on days 10, 13, and 16 following implantations. (B) Survival curves for mice in the different treatment groups. Statistical analysis was performed by Log-rank (Mantel-Cox) test, * P < 0.05, ** P < 0.01, ns- non significant. (C) Ultrasound images of orthotopic tumors eight days after treatment. (D) Quantification of KPC tumor sizes eight days after treatment. Statistical analysis was performed using one-way ANOVA and multiple comparisons using Holm-Sidak, * P < 0.05, ** P < 0.01. (E) Timeline for the biodistribution experiment in mice with orthotopic human DAOY medulloblastoma tumors. (F) Live IVIS imaging of mice treated with fluorescent TMAB3 over the course of eight days. (G) Timeline for the efficacy experiment in mice with DAOY tumors. Mice were treated IV with a single dose of PBS, 3p-hpRNA, TMAB3, temozolomide, or TMAB3/3p-hpRNA complexes. (H) Live bioluminescent images of mice bearing orthotopic medulloblastoma tumors ten days after treatment by IVIS imaging. (I) Quantification of tumor burden as measured by bioluminescence. Data from n=4 mice per group. One-way ANOVA and multiple comparisons using Holm-Sidak, * P < 0.05, ** P < 0.01. (J) Timeline for efficacy experiment in mice bearing B16.F10-OVA flank tumors. Mice were treated with three doses of IV PBS, TMAB3, 3p-hpRNA, or TMAB3/3p-hpRNA. (K) Representative images of tumors in treated mice taken at day 24 post-implantation. (L) Tumor growth curves following the indicated treatments. Data from n=5–6 mice per group. One-way ANOVA of repeated measurements with multiple comparison by Holm-Sidak, * P < 0.05, ** P < 0.01, **** P < 0.0001 (M) Body weights of treated mice bearing B16.F10-OVA flank tumors over the course of the experiment. Data from n=5–6 mice per group with statistical analysis performed by One-way ANOVA.

Fig. 4.. TMAB3/3p-hpRNA treatment increases the activation of CD8⁺ T cells in the PDAC microenvironment.

(A) Schematic of mouse survival experiment to assess the efficacy of TMAB3/3p-hpRNA treatment in mice bearing orthotopic KPC tumors, compared to controls. Mice were treated IV with three doses of: TMAB3/3p-hpRNA complexes, 3p-hpRNA, TMAB3, TMAB3/inert siRNA, and PBS, on days 10, 13, and 16 following tumor implantations and followed daily by observation. (B) Survival curves following the different treatments. n=7–8 mice per group with statistical analysis performed by Log-rank (Mantel-Cox) test, * P < 0.05, ** P < 0.01. Survival endpoints were defined based on the approved animal use protocol (details in Materials and Methods). (C) Representative flow cytometry analysis of CD3⁺ and CD8a⁺ T cells in pancreatic tumors from mice treated as in panel A and harvested 7 days after the 3rd dose for flow cytometry. (D and E) Quantification of the percentages of CD8⁺ T cells out of total CD45⁺ cells at 1 day (D) and 7 days (E) after the 3^rd treatment, based on flow cytometry (n = 7 biologically independent samples, with mean and SEM shown). One-way ANOVA with multiple comparison by Holm-Sidak, **** P < 0.0001. (F and G) Immunoblot analyses of CD8 expression in lysates of pancreatic tumors harvested 1 day and 7 days after the 3^rd treatment, as indicated. Quantification of CD8 expression was normalized to the β-actin loading control using ImageJ (n=3 biological replicates, with mean and SEM shown). One-way ANOVA with multiple comparison by Holm-Sidak, **** P < 0.0001. (H) Analysis of activated CD8⁺ T cells identified by expression of granzyme B by flow cytometry, 7 days after treatment. (I) Quantification of percentage of activated CD8+ T cells as measured by granzyme B. One-way ANOVA with multiple comparison by Holm-Sidak, *** P < 0.001 (J) Flow cytometry analysis of cells expressing CXCR6 and CD8 within mouse pancreatic tumors. (K) Graph summarizing percentage of CXCR6⁺ resident memory CD8⁺ T cells in tumors of mice in the different treatment groups. One-way ANOVA with multiple comparison by Holm-Sidak, *** P < 0.001. (L) Timeline to assess the role of lymphocytes in the anti-tumor efficacy of TMAB3/3p-hpRNA complexes by comparing efficacy in Rag1 KO mice bearing KPC orthotopic tumors to wild-type tumor-bearing mice. Mice were treated with three IV doses of: TMAB3/3p-hpRNA complexes, 3p-hpRNA, TMAB3, TMAB3/inert siRNA, and PBS, on days 10, 13, and 16 following tumor implantations. (M) Survival following treatments in Rag1 KO mice (n=5). The survival curve for wild-type mice is shown as the dashed line and is derived from Fig. 4B, n=8. Statistical analysis was performed by Log-rank (Mantel-Cox) test.

Fig. 5.. Single-cell RNA sequencing of tumors reveals increased activation of T cells following TMAB3/3p-hpRNA treatment.

(A) Schematic of scRNA-seq study design. Mice bearing orthotopic KPC tumors were treated with three IV doses of: TMAB3/3p-hpRNA complexes, TMAB3/siRNA TMAB3 alone, or PBS on days 10, 13, and 16 following implantation. On day 23, tumors were harvested and analyzed by scRNA-seq. (B) UMAP analysis of single-cell RNA-seq for all cell lineages present in the tumor. (C) Bubble plot of the genes used to identify each cell lineage (as shown in the UMAP). (D) Bar graph quantification of the identified immune-cell subtypes. (E to G) Bubble plots of: (E) RIG-I stimulated genes; (F) type I IFN stimulated genes; (G), Extrinsic apoptosis genes in cancer cells following the indicated in vivo treatments. (H and I) Bubble plots of activation (H) and exhaustion (I) markers in CD8⁺ T cells. (J) Schematic of co-culture assay design. KPC wild type, RIG-I or ENT2 knockout (KO) KPC cells were seeded into assay plates the next day treated with TMAB3/3p-hpRNA complexes, 3p-hpRNA, TMAB3, or PBS. Six hours later, mouse T cells were added to the culture. After 48 hours of incubation, ELISpots were conducted for the indicated cytokines. (K and L) Immunoblot visualization to confirm absence of protein in KPC RIG-I KO (K) and ENT2 KO (L) cells. B-actin was used as a loading control. (M to O) ELISpot results for IFNγ (Μ), TNFα (Ν), and IL-10 (O) in the co-culture assays. One-way ANOVA with multiple comparison by Holm-Sidak, **** P < 0.0001; ns- not significant.